Germany: Published Guideline AI Security Concerns

On 14 April 2023, the German Federal Office for Information Security published a guideline titled AI Security Concerns in a Nutshell, which outlines the most relevant attacks on machine learning systems and potential appropriate defences. The guideline states that, as machine learning is increasingly used in critical areas like autonomous driving, cancer detection, and biometric authentication, the risk of attacks escalates. Specifically, the guideline provides details about evasion attacks, information extraction attacks, and poisoning and backdoor attacks, and suggests appropriate defences.